Note Effects of drought on the induced defence reaction of Scots pine to bark beetle-associated fungi L Croisé F Lieutier INRA, Station de Zoologie Forestière, Ardon, 45160 Olivet, France (Received 3 April 1992; accepted 14 September 1992) Summary — Water stress was imposed on Scots pine saplings grown in a greenhouse. Predawn leaf water potential (ψ wD ) was monitored regularly while carbohydrate reserves were quantified in the phloem, xylem and shoots. Trees were inoculated in the bole with a bark beetle-associated fun- gus at 4 different periods of the year; the induced defence reaction was quantified 2-3 wk later by its length and the total amount of resin. A decrease in the induced reaction (length and resin quantity) was observed in the most severely stressed trees. However, no clear relationship could be estab- lished between the concentration of carbohydrates (soluble or hydrolysable) and the intensity of the defence reaction. pinus sylvestris / phytopathogenic fungus / induced defence reaction / water stress / pre- dawn leaf water potential / carbohydrate reserve Résumé — Action d’un stress hydrique sur la réaction de défense induite du pin sylvestre contre des champignons issus d’insectes Scolytides. Un stress hydrique a été appliqué en serre sur des pins sylvestres de 6 ans. Des champignons phytopathogènes préalablement isolés d’in- sectes Scolytides ont été directement inoculés dans le liber de ces arbres afin de préciser l’effet de la contrainte hydrique sur les caractéristiques de la réaction locale de défense (longueur et quantité totale de résine). L’état hydrique des arbres a été suivi régulièrement par des mesures du potentiel hydrique de base au niveau des aiguilles (ψ wD). Les réserves glucidiques de l’arbre ont été dosées dans le liber, l’aubier et les pousses au moment des inoculations. Quatre séries d’expérience ont ainsi été réalisées sur les mêmes arbres, de mai à octobre 1989. Pour des potentiels de base allant jusqu’à -2 MPa, ψ wD est systématiquement et négativement corrélé avec les caractéristiques de la réaction de défense dans chaque série d’expériences (tableau I). Aucune différence concernant la teneur en glucides n’apparaît entre les arbres stressés et les témoins dans les différents comparti- ments considérés (tableau II). Pinus sylvestris / champignon phytopathogène / réaction de défense induite / stress hydrique / potentiel hydrique foliaire / réserve glucidique INTRODUCTION In coniferous trees attacked by bark beetles, an induced reaction visible as a resinous impregnation of the tissues sur- rounding the point of aggression plays a determining role in the resistance of the tree. In the majority of cases, the reac- tion appears to be induced, or at least considerably amplified by the presence of fungi introduced by the insect (Berry- man, 1972; Christiansen and Horntvedt, 1983; Raffa and Berryman, 1983; Cook and Hain, 1986; Christiansen et al, 1987; Lieutier et al, 1988; Lieutier, 1992). The success or failure of each attack de- pends on the outcome of the struggle be- tween the bark beetle and its associated fungi, and the intensity of the defence re- action at the site of attack. The defen- sive ab,lity of a tree can be represented by the attack density threshold above which the tree is no longer able to resist. Since the production of induced resin is costly in terms of energy (Croteau and Loomis, 1975) this threshold should be higher the lower the energy mobilized in each reaction. The attack density thresh- old appears to depend on the vigor of the tree at the time of attack (Berrymn, 1978; Waring and Pitman, 1983; Mulock and Christiansen, 1986). Tree vigor is conditioned by several factors among which climatic factors are probably deter- mining. A number of authors have report- ed that the worst bark beetle damage in conifer forests are often preceded by a major period of drought (Christiansen et al, 1987; Mattson and Haack, 1987; and references therein). Thus, there is prob- ably a relationship between the water status of a tree, the attack density threshold, and the characteristics of the induced reaction. The present paper in- vestigates the relationships between tree water status and the characteristics of the induced reaction. MATERIALS AND METHODS Water stress treatments Fourteen 6-yr-old (2.20 rri high) Scots pines grown in pots were divided into 2 groups and watered with a drop-by-drop system. Group A (control) received = 20 I of water per wk per tree, while group B was subjected to water stress be- ginning on May 12, Julian d 133. After this date, the group B trees received 3 I of water per wk per tree until June 16, Julian d 168, 9 I from June 16-July 3, Julian d 185, 6 I from July 3- August 7, Julian d 220, and again 3 I per wk per tree thereafter. Tree water status Predawn leaf water potential (ψ wD ) was meas- ured with the pressure chamber technique (Scholander et al, 1965). The measurements were carried out on 2 pairs of needles per tree from the current year shoots. Needles were col- lected between 4:30 and 5:30 am GMT every 10-20 d. Stress intensity at a given date of inoc- ulation was evaluated as the average (ψ wD ) of measurements made on the same tree between 6 d prior to and 16 d after inoculation. Tree defence reactions The reactions were induced with artificial inocu- lations of Ophiostoma brunneo-ciliatum (Math- K) or Leptographium wingfieldii (Morelet), 2 As- comycete fungi previously isolated from the pine bark beetle Ips sexdentatus Boern and Tomicus piniperda L respectively (Lieutier et al, 1989). These fungi had been cultivated and purified as monospore cultures on a malt-agar medium. Five-mm diameter agar implants from 3-wk-old cultures were introduced into the trees at the cambium level accoring to a technique derived from Wright (1933) and which has been previ- ously described (Lieutier et al, 1989). One inocu- lation was performed per tree on May 22 (Julian d 143), July 11 (Julian d 193), 1989 with O brun- neo-ciliatum, August 16 (Julian d 229), and Sep- tember 21 1989 (Julian d 265) with L wingfieldii. Two or 3 wk after inoculation, the bark around the inoculation points was pulled away to ex- pose the reaction zone of the phloem. This reac- tion zone was then measured (length) and cut into samples which were immediately placed into dry ice under a nitrogen atmosphere. These samples were kept at -35 °C until analysis for total resin content. Samples of non-inoculated phloem, as well as sapwood and shoot axes (phloem and xylem) of new growth were taken on December 21, 1988, March 20, 1989, and on the days of inoculation; they were frozen, and stored in the laboratory at -35 °C before analy- sis for their carbohydrate content. Non- inoculated phloem and sapwood were sampled near the inoculation point, and the shoots at the end of lateral branches directly above the inocu- lation point. Analyses The resin content in the induced reaction zones was measured according to a previously de- scribed method (Lieutier et al, 1989). The re- sults were expressed as the total quantity of res- in present in fresh whole reaction zones. Carbohydrates, separated into a soluble and hy- drolysable fraction, were measured and ex- pressed in terms of glucose equivalents by the anthrone colorimetric method used by Mokrash (1954) and modified by Sauvard (1988). RESULTS Water status of trees Watering of control trees kept their pre- dawn leaf water potential at relative stable levels between -0.3 and -0.6 MPa during the whole experiment (fig 1). In stressed trees, 2 phases can be distinguished. The first coincides with the first inoculations and was characterized by a strong water deficit (ψ wD down to -2.1 MPa). The sec- ond, where ψ wD was between -0.55 and -1.2 MPa, began at Julian d 182 and con- tinued until the end of the expriment. Defence reactions The defence reaction zone of the stressed trees was only once significantly shorter than that of the control trees (May; 15 ± 4.7 mm for the control and 5.6 ± 3.0 mm for the stressed trees). Total quantity of resin in the defence reaction zone never differed significantly between the 2 categories of trees. However, the correlations between |ψ wD | and the characteristics of the de- fence reactions (length and resin quantity) were constantly negative (table I). Three of them (2 for the length and 1 for resin quan- tity) differed significantly (P ≤ 0.05) from zero and another (concerning resin quanti- ty) was almost significant (P = 0.058). Carbohydrate reserves (table II) After statistical analyses, phloem, xylem and shoots showed no significant differ- ences between the carbohydrate (soluble or hydrolysable) content of stressed trees and that of control trees. Otherwise, the carbohydrate content of each compartment did not vary very much over time. DISCUSSION Since the trees recovered from all inocula- tions it may be hypothesized that the length of the reaction zone was proportion- al to the duration of the struggle between the tree and its aggressor, and that total quantity of induced resin included in the re- action zone was proportional to the quanti- ty of energy invested by the tree in its de- fence. Under these conditions, our results suggest that the aggressors were arrested more rapidly and that the quantity of ener- gy invested for that purpose was lower in the stressed trees. We cannot a priori discard a direct ef- fect of stress on the fungus. However, phy- topathogenic fungi are generally more tol- erant than plants to water deficits (Pinon, 1986; and references therein). According to Cook and Papendick (1972; in Pinon, 1986), fungal growth is still possible for wa- ter potentials as low as -3 to -5 MPa. Sometimes, water stress can even pro- voke stimulation of that growth, as ob- served by Bagga and Smally (1967; in Pin- on, 1986) for the aspen canker in in vitro cultures. Under these conditions it seems unlikely that the decrease in the induced reaction in our experiment was due to a decrease in fungal vitality. Some data on response regarding the relationships between water stress and the induced defence reaction have been ob- tained. A constant negative correlation was observed between the characteristics of the defence reaction and the predawn leaf water potential. Thus, it seemed that when water stress increased, both reaction length and quantity of induced resin in that reaction decreased. The decrease in the amount of induced resin is in agreement with the decrease observed by Lorio (1986) for constitutive resin in Pinus taeda during a period of severe drought. It is also in agreement with the findings of Paine and Stephen (1987) who noticed for the same species a less important induced re- action in the dominated trees than in the dominant trees. Stephen et al (1983) claimed that an important defence reaction would correspond to trees resistant to bark beetle attacks. In Norway spruce, Chris- tiansen et al (1987) observed that marked reactions could develop in weak trees close to death. This observation seems to be in opposition to our present results; however the situation described by these authors was obtained with an inoculation density above the lethal threshold; in these conditions, the fungus extended to the whole phloem. It was very different from our situation with isolated inoculations which were always contained by the tree response. Nevertheless, Lorio (personal communication) observed an increase in constitutive resins during a limited period of drought, and Lieutier and Ferrell (1988) reported an increase of induced reaction in Scots pine when tree growth efficiency de- creased. Our results did not demonstrate any modification in the amount of stored carbo- hydrates in the shoots, phloem and xylem in response to water stress (table II). Pos- sibly a longer or a more intense water stress would have been necessary to in- duce such modifications. Indeed, Grieu et al (1988) reported an increase of the solu- ble carbohydrates in Douglas fir needles for strong water stress (ψ wD = -1.6 MPa). However, these authors observed only weak variations of these compounds in the needles and in the roots of the same tree and in those of Pseudotsuga macrocarpa and Cedrus atlantica, with a leaf predawn water potential below -2 MPa. It is there- fore difficult to consider the observed varia- tions in the tree defence reactions to be a consequence of variation of stored carbo- hydrates. These conclusions agree with the re- sults of Christiansen and Ericsson (1986) who reported that the level of stored starch was not correlated with Picea abies resis- tance to infection by Ophiostoma poloni- cum. Owing to the fact that resin synthesis is costly for the tree (Croteau and Loomis, 1975), Christiansen and Ericsson (1986) have suggested that the flow of assimilates might be the main source of energy for the development of the defence reactions. Stored starch might thus be a complemen- tary source of energy when the current flow is not sufficient. Otherwise, the study of resin biosynthesis in Pinus pinaster (Bernard-Dagan, 1988) suggested that the ascending flux of soluble carbohydrates from roots might also be an important en- ergy source for the induced reactions. Ac- cording to the same authors, the possible catabolism of the heartwood resin might lo- cally participate in the defence mecha- nisms at the beginning of the annual growth activity. ACKNOWLEDGMENTS The authors are grateful to E Christiansen (NISK, Norway) and E Dreyer (INRA, France) for their comments on the manuscript. They also thank J Garcia and P Romary for their tech- nical help. REFERENCES Bernard-Dagan C (1988) Les substances de ré- serve du Pin maritime: rôle éventuel des métabolites secondaires. Bull Soc Bot Fr Ac- tual Bot 135, 25-40 Berryman AA (1972) Resistance of conifers to invasion by bark beetle-fungus associations. Bio-Science 22, 599-601 Berryman AA (1978) A synoptic model of the Lodgepole pine/Mountain pine beetle interac- tion and its potential application in forest management. In: Theory and Practice of Mountain Pine Beetle Managment in Lodge- pole Pine Forests (Berryman AA, Amman GD, Stark RW, Kibbee DL, eds) College For Res, Univ Idaho, Moscow, ID, 98-105 Christiansen E, Horntvedt R (1983) Combined Ips/Ceratocystis attack on Norway spruce, and defensive mechanisms of the trees. Z Angew Entomol 96, 110-118 Christiansen E, Ericsson A (1986) Starch re- serves in Picea abies in relation to defence reaction against a bark beetle transmitted blue-stain fungus, Ceratocystis polonica. Can J For Res 16, 78-83 Christiansen E, Waring RH, Berryman AA (1987) Resistance of conifers to bark beetle attack: searching for general relationships. For Ecol Manage 22, 89-106 Cook SP, Hain FP (1986) Defensive mecha- nisms of Loblolly and Shortleaf pine against attack by southern pine beetle, Dendrocto- nus frontalis Zimmerman, and its fungal as- sociate, Ceratocystis minor (Hedgecock) Hunt. J Chem Ecol 12, 1397-1406 Croteau R, Loomis WD (1975) Biosynthesis and metabolism of monoterpenes. Int Flavours Food Addit 6, 292-296 Grieu P, Aussenac G, Larher F (1988) Séche- resse édaphique et concentrations en quelques solutés organiques des tissus foli- aires et racinaires de trois espèces de co- nifères: Cedrus atlantica Manetti, Pseudotsu- ga macrocarpa (Torr) Mayr, Pseudotsuga menziesii (Mirb) Franco. Ann Sci For 45, 311-322 Lieutier F (1992) Induced defence reaction of conifers to bark beetles and their associated Ophiostoma. In: Ceratocystis and Ophiosto- ma: Toxonomy, Biology and Pathogenicity (Wingfield MJ, Seifert KA, Webber JF, eds) Am Phytopathol Soc Press, 206-215 Lieutier F, Ferrell GT (1988) Relationships be- tween indexes of tree vigour and the induced defense reaction of Scots pine to a fungus associated with Ips sexdentatus (Coleoptera: Scolytidae). In: Integrated Control of Scolytid Bark Beetles (Payne TL, Saarenmaa H, eds) Virginia Polytechnic Inst and State Univ, Blacksburg, VI, 163-178 Lieutier F, Yart A, Garcia J, Poupinel B, Levieux J (1988) Do fungi influence the establishment of bark beetles in Scots pine? In: Mechanisms of Woody Plant Defenses Against Insects; Search for Pattern (Mattson WJ, Levieux J, Bernard- Dagan C, eds) Springer Verlag, 321-331 Lieutier F, Yart A, Garcia J, Ham MC, Morelet M, Levieux J (1989) Champignons phytopa- thogènes associés à deux coléoptères Scoly- tidae du Pin sylvestre (Pinus sylvestris L) et étude préliminaire de leur agressivité envers l’hôte. Ann Sci For 46, 201-216 Lorio PL Jr (1986) Growth-differentiation bal- ance: a basis for understanding Southern pine beetle-tree interaction. For Ecol Man- age 14, 259-273 Mattson WJ, Haack RA (1987) The role of drought in outbreaks of plant-eating insects. Bio Science 37, 110-118 Mokrash LC (1954) Analysis of hexose phos- phates and sugar mixtures with the anthrone reagent. J Biol Chem 208, 55-59 Mulock P, Christiansen E (1986) The threshold of successful attack by Ips typographus on Picea abies: a field experiment. For Ecol Manage 14, 125-132 Paine TD, Stephen FM (1987) Influence of tree stress and site quality on the induced de- fense system of Loblolly pine. Can J For Res 17, 569-571 Pinon J (1986) Les maladies de faiblesse en plantation. Rev For Fr 38, 324-330 Raffa KF, Berryman AA (1983) Physiological as- pects of Lodgepole pine wound responses to a fungal symbiont of the mountain pine bee- tle, Dendroctonus ponderosae (Coleoptera: Scolytidae). Can Entomol 115, 723-731 Sauvard D (1988) Capacité de multiplication de Tomicus piniperda L (Coleoptera: Scolytidae) selon la densité d’attaque et diverses carac- téristiques de l’arbre. Thèse Doctorat, Univer- sité d’Orléans La Source, 130 p Scholander PF, Hammel HT, Bradstreet ED, Hemmingsen EA (1965) Sap pressure in vas- cular plants. Science 148, 339-346 Stephen FM, Paine TD, Lih MP (1983) Under- standing bark beetle/host interactions: a means for improving decision strategies. Z Ang Entomol 96, 257-265 Waring RH, Pitman GB (1983) Physiological stress in Lodgepole pine as a precursor for Mountain pine beetle attack. Z Ang Entomol 96, 265-270 Wright E (1933) A cork-borer method for inocu- lating trees. Phytopathology 23, 487-488 . and con- tinued until the end of the expriment. Defence reactions The defence reaction zone of the stressed trees was only once significantly shorter than that of the control. difficult to consider the observed varia- tions in the tree defence reactions to be a consequence of variation of stored carbo- hydrates. These conclusions agree with the re- sults. Note Effects of drought on the induced defence reaction of Scots pine to bark beetle-associated fungi L Croisé F Lieutier INRA, Station de Zoologie Forestière, Ardon, 45160 Olivet,